Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

This is exercise 70, chapter 4. from Folland (page 142)

Let $X$ be a compact Hausdorff space. An ideal in $C(X, \mathbb{R})$ is a subalgebra $J$ of $C(X, \mathbb{R})$ such that if $f\in J$ and $g\in C(X, \mathbb{R})$ then $fg\in J$.

  1. If $J$ is an ideal in $C(X, \mathbb{R})$, let $h(J) = \{x \in X: f(x) = 0,\ \forall f \in J\}$. Then $h(J)$ is a closed subset of $X$, called the hull of $J$.
  2. If $E\subset X$, let $k(E)=\{f \in C(X, \mathbb{R}) : f(x)=0,\ \forall x \in E\}$. Then $k(E)$ is a closed ideal in $C(X, \mathbb{R})$, called the kernel of $E$.
  3. If $E\subset X$, then $h(k(E)) =\overline{E}$.
  4. If $J$ is an ideal in $C(X, \mathbb{R})$ then $k(h(J))=\overline{J}$. (Hint: $k(h(J))$ may be identified with a subalgebra of $C_0(U, \mathbb{R})$ where $U=X\setminus h(J)$)

I've managed to prove assignments 1-3. In (4)

if $f$ is from $J$ then for each $y\in h(J)$ is $f(y)=0$, then must be $f\in k(h(J))$. Since $k(h(J))$ is closed, also $\overline J\subset k(h(J))$.

For other way around, I've proven the hint ($k(h(J))$ may be identified with a subalgebra of $C_0(U, \mathbb{R})$ where $U=X\setminus h(J)$), using some corollary fo Stone-Weierstrass), but don't understand how to make conection between $C_0(X\setminus h(J), \mathbb{R})$ and $\overline{J}$


P.S. This is my first time asking question, but I've found this site very helpfull, so thanks!

share|cite|improve this question

I don't know for the moment how to use the hint. However, we can try the following.

  1. Fix $f\in k(h(J))$ and $\varepsilon>0$. Let $K:=\{x\in X, |f(x)|\geqslant \varepsilon\}$. By compactness, we can find $g\in J$ non-negative such that $g>0$ on $K$. Indeed, $K$ is compact, and the intersection $\bigcap_{g\in J}g^{-1}(\{0\})\cap K$ is empty, so there are $g_1,\dots,g_n$ such that $K\cap\bigcap_{j=1}^ng_j^{-1}(\{0\})\cap K$ is empty. Then take $g:=\sum_{j=1}^ng_j^2$.

  2. Let $f_n(x):=\frac{f(x)}{\frac 1n+g(x)}g(x)$. It's an element of $J$ such that $\lVert f_n-f\rVert\leqslant 2\varepsilon$.

It's also exercise 1.5.b) in Elements of Functional Analysis, by F. Hirsch and G. Lacombe, Springer

share|cite|improve this answer
First, forgive me for taking so long to come to answer. I can't figure out why function g exists. – Ana Apr 11 '13 at 10:34
@ana I've added details. – Davide Giraudo Apr 11 '13 at 11:38

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.